Cryptogenic Stroke: The role of silent Atrial Fibrillation

Size: px

Start display at page:

Download "Cryptogenic Stroke: The role of silent Atrial Fibrillation"

Sherilyn Olivia Lawrence
5 years ago
Views:

1 Cryptogenic Stroke: The role of silent Atrial Fibrillation PD Dr. med. Martin A. Ritter Westfälische Wilhelms-Universität Münster Germany 1 Turin,

2 Agenda 1. What is a cryptogenic? stroke Understanding the burden of cryptogenic stroke for individual patients and for society 2. Why bother with paroxysmal Atrial Fibrillation (AF)? 3. How to reliably detect (rare episodes of) paroxysmal AF? Strategies Can we predict paroxysmal AF? 4. What to do then? discussion 2 Turin,

Cardio-embolic stroke What is a cryptogenic stroke?

3 Cardio-embolic stroke What is a cryptogenic stroke? Stroke classification according to TOAST Atrial fibrillation, criteria Ventricular aneurysm / thrombus etc. ~30% atherothrombosis Lacunar stroke Carotid stenosis / intracranial stenosis > 50% Aortic plaques, instable carotid plaque? ~15% Cerebral microangiopathy (Hypertension, Diabetes) ~15% Other defined etiologies Periprocedural, dissections, vasculitis, PFO, etc. ~10% Undetermined etiology Concurring etiologies, incomplete work-up cryptogenic strokes ~30% 3 Turin, Adams HP Jr, et. Al. Stroke. 1993;24:35-41.

4 What is a cryptogenic stroke? And understanding the burden of it A cryptogenic stroke is a stroke of unknown cause. The diagnosis demands a thorough and complete work-up of embolic or thrombotic sources 20-30% of all strokes have to be labeled as cryptogenic. That is about Germans or Italians each year! The diagnosis leaves doctors and patients unsure of the risk of future stroke and optimal secondary prevention big psychological problem! We suspect that a number of these patients does have AF we did not detect. 4 Turin,

5 Why bother with paroxysmal atrial fibrillation? Because risk of stroke does not differ greatly between paroxysmal or permanent AF! 1981 patients, Follow-up: 3.6 years Incidence rate of ischemic stroke: Cumulative 2,6% (paf) vs. 2,9% (caf) per year 5 Turin, Friberg L, et al. Eur Heart Journal 2009

6 How to track (rare episodes of) paroxysmal AF? Simple answer: Increase the time of monitoring! Do Holter ECG! Repeat Holter Do telemetry while in hospital! Do 3-7 day Holter Use external loop recorder (MCOT) Do continuous Monitoring (implantable cardiac monitors, ICM) 6 Turin,

7 Study overview: Detection rates of iaf with extended monitoring methods Definition of iaf varied widely In many studies episodes <30s of AF were accepted, and accounted for >50% of all pt.! 7 Turin, Khan M, Miller DJ. Detection of paroxysmal atrial fibrillation in stroke/tia patients. Stroke Res Treat. 2013;2013:

Increased monitoring times after stroke result in higher yields of silent AF! Stahrenberg R. Stroke. 2010; 41: 2884-2888. Gladstone DJ. NEJM.

8 Increased monitoring times after stroke result in higher yields of silent AF! Stahrenberg R. Stroke. 2010; 41: Gladstone DJ. NEJM. 2014; 370;26: 2467 Only 69% of patients tolerated the device 5 days or more! Only 4 out of 5 patients had data for 21 days or more! 8 Turin,

9 Prolonged Monitoring = higher Yield Comparison studies in stroke patients Autor n Methodik / Dauer Yield Stahrenberg Stroke d Holter vs. 24h Holter 12,5 % vs. 4,8% Grond Stroke h Holter vs. 24h Holter 4,3% vs. 2,9% Rizios Stroke h stroke unit monitoring (dedicated analysis tool) vs. 24h Holter 7,6% vs. 2,8% Gladstone NEJM d externer loop recorder vs. 24h Holter 16,1% vs. 2,2% Ritter Stroke ICM (Reveal XT ) vs. 7d Holter 17,1% vs. 1,6% 9 Turin,

10 Continuous monitoring has the highest diagnostic yield Implantable cardiac monitors (ICMs) should have the highest yield 10 Turin, Botto G, et al. J Cardiovasc Electrophysiol 2009; 20:

ICMs are implanted subcutaneously under local

11 ICMs are implanted subcutaneously under local anesthesia e.g.: Reveal XT algorithm detects episodes >2 min. (programmable) XPECT study, : sensitivity (96,1%) and specificity (85,4%) vs. Cardiologist evaluated Holter monitoring in pt. c. known iaf Hindricks G, et al.. Circ Arrhythm Electrophysiol Turin,

12 Pilotstudies using ICMs in cryptogenic stroke patients Autor n Duration of monitoring Median until first episode of iaf Additional Yield Ritter Stroke Median 397 d 64 d 17.1% Cotter Neurology Median 229 d (withoutaf) 48 d 25.5% Etgen Stroke year 153 d 27.3% Christensen Eur J Neurol Mean 569 d Mean 109 (±48) d 16.1% 12 Turin,

13 CRYSTAL-AF, n= Turin, Sanna T., et al: N Engl J Med 2014;370:

CRYSTAL AF: Baseline Characteristics: ICM Control Age 61.6 ± 11.4 years 61.4 ± 11.3 years Gender - Male 142 (64.3%) 138 (62.7%) Index Event Stroke 200 (90.5%) 201 (91.4%) Index Event TIA 21 (9.

14 CRYSTAL AF: Baseline Characteristics: ICM Control Age 61.6 ± 11.4 years 61.4 ± 11.3 years Gender - Male 142 (64.3%) 138 (62.7%) Index Event Stroke 200 (90.5%) 201 (91.4%) Index Event TIA 21 (9.5%) 19 (8.6%) Pre-enrollment AF screening Holter Monitoring Pre-enrollment AF screening Telemetry Time between index event and randomization Time to randomization and device insertion 71.5% of patients Median of 23 hours (IQR 21-24) 29.9% of patients Median of 48 hours (IQR 36-96) 70.9% of patients Median of 24 hours (IQR 22-24) 29.5% of patients Median of 72 hours (IQR 48-96) 36.6 ±28.2 days 39.6 ±26.9 days 8.7 ± 27.6 days n/a 14 Turin,

15 CRYSTAL AF, Secondary Endpoint: Detection of AF at 12 months 12,4% 2,0% 15 Turin, Sanna T., et al: N Engl J Med 2014;370:

16 Can we predict iaf? Can we select the ideal patient for prolonged monitoring? - Clinical predictors: - age - CHADS 2 andcha 2 DS 2 -Vasc Score - Echoparameters: - E.g. atrial size - ECG parameters - e.g. SPBs 16 Turin,

17 Predict iaf by clinical scores? no significant clinical predictors in our study noaf (n=50) iaf (n=10) age 61; ; n.s. Female % (n) 40 (20) 60 (6) n.s. Hypertension % (n) 70 (35) 70 (7) n.s. Diabetes % (n) 14 (7) 0 p=0.1 CHF % (n) 0 0 n.s. Vascular disease % (n) 8 (4) 0 n.s. CHADS 2 -Score 3; 2-3 3; n.s. CHA 2 DS 2 -VASC Score 4; 3-5 4; 3-5 n.s. Insular cortex involvement 14 (7) 10 (1) n.s. Total monitoring time[d] Time to implantation after event [d] 397; ; ; ; n.s. n.s. 17 Turin, Ritter MA et al.; Stroke 2013

18 Cotter et al.: Classical risk factors predict iaf Cotter PE, et al. Incidence of atrial fibrillation detected by implantable loop recorders in unexplained stroke. Neurology Apr 23;80(17): Turin,

19 Prediction of iaf according to purely clinical risk scores is ineffiecient! 1368 patients screened 416 had AF within 1,1 y. No differences in CHADS 2 - score However: Different population (TRENDS study, known cardiac disease) Different devices (ICD, PM, CRT) 19 Turin, Ziegler PD et al. Am J Cardiol. 2012:110:

20 Subgroup Analysis of CRYSTAL AF: Trend for higher yield of iaf in higher CHADS2 scores 20 Turin,

21 Cotter et al.: Classical risk factors predict iaf - and also echo and ECG parameters! Cotter PE, et al. Incidence of atrial fibrillation detected by implantable loop recorders in unexplained stroke. Neurology Apr 23;80(17): Turin,

22 22 Turin,

23 Thoughts on consequences for future patient selection for ICMs - 1 ILRs are much more sensitive than 7 day Holter! Most patients detected within 3 mo. from implantation, but there is significant yield beyond this! 7 or 30 days of monitoring are not enough! Emergent questions: How long should a monitoring be performed (in order to attribute a positive result to the recent event)? Is anticoagulation really the answer for all patients with iaf ( AF burden ) How about PFO and iaf? 23 Turin,

24 Thoughts on consequences for future patient selection for ICMs 2 Who is a candidate for extensive monitoring = ICM? Complete work-up performed and stroke is cryptogenic Patients with numerous SBPs and SVRs on 24h Holter You think it s AF (stroke pattern plus risk factors but you can not prove it yet!) Relatively fit patients Compliance with telemetry a lot to lose 24 Turin,

25 Summary % of patients with cryptogenic stroke do have iaf 2. Detection requires at least 3-12 months extensive monitoring 3. It is not possible to predict the occurrence of AF according to clinical risk factors the final word on paraclinical parameters (echo, 24h Holter) has not been spoken 4. Up to now any diagnosis of AF according to cardiological guidelines is currently relevant and should be followed by anticoagulaton 25 Turin,

26 Thank you for your attention! 26 Turin,

ECG monitoring after ischemic stroke of TIA of unknown source with an insertable monitor? YES

ECG monitoring after ischemic stroke of TIA of unknown source with an insertable monitor? YES Isabelle C Van Gelder University Medical Center Groningen The Netherlands ESC stroke council Prague January